Intrusion detection system for cloud forensics using bayesian fuzzy clustering and optimization based SVNN

Abstract

Intrusion detection has emerged as one of the major challenges involved in the cloud forensics. This work introduces an intrusion detection framework for the cloud environment with clustering and two-level classifiers. In the first step of the process, a Bayesian fuzzy clustering is used for clustering the nodes in the cloud. And in the next step, two-level gravitational group search-based support vector neural network (GG-SVNN) classifier identifies intrusion in clusters. GG-SVNN is a novel optimization scheme proposed in this work, by combining the group search optimizer, and gravitational search algorithm. The intrusion information provided by level 1 classifier is arranged to form compact data, and provided to the level 2 classifier. The level 2 classifier finally identifies total nodes affected by the intruders. The simulation of the proposed intrusion detection is done with the help of KDD cup dataset. From the simulation results, it is evident that the proposed GG-SVNN classifier has achieved overall best performance by achieving high accuracy value of 92.41% and low false alarm rate of 4.75% respectively.